Baker Fund Proposal Checklist

Applicants must complete and sign the checklist. The checklist should be included as the second page of the application (following the cover page). Cover page use Baker form Checklist use Baker form Abstract* 1 double-spaced page Introduction (for continuations or resubmissions only)* 1 double-spaced page Discussion 10 double-spaced pages Glossary/Definition of Terms* (not required) 2 double-spaced pages Bibliography (not required) 3 pages Biographical Information (applicant(s) and key personnel) 3 pages per person Other Support (applicant(s) and key personnel) 1 page per person Budget and Justification no limit specified Appended Materials 10 pages; no more than 10 minutes of footage Recommended Reviewers 5 required Electronic copy of proposal Single Acrobat file, containing

entire proposal and required signatures

* These sections should be written in language understandable by an informed layperson to assist the committee in its review. **Please note: The committee has the right to return without review any proposals that do not conform to these format requirements.** Applicant signature: __________________________________________________________________

✔

✔

✔

✔

✔

✔

✔

✔

✔

✔

Hooper Baker

1

Abstract

Nervous systems create behavior. Neurons, the cells of the nervous system, show a wide range of

electrical activities even in isolation. The observation indicates that different neurons are

inherently different. The most important determinants of neuron inherent electrical activity are

voltage-dependent conductances, proteins in neuron membranes that open and close when the

voltage across the membrane changes. The different electrical activities of different neurons thus

likely arise in part from the neurons having different sets of voltage-dependent conductances.

One might assume that a direct correspondence can be made: activity A stems from voltage-

dependent conductance set A; activity B from set B, etc. The difficulty with this hypothesis is

that changes in one voltage dependent conductance can be compensated for by changes in other

voltage dependent conductances. Multiple voltage conductance sets can therefore produce

essentially identical neuron activity. Resolving this redundancy problem is difficult because 1)

present techniques allow measuring, in any one experiment (that is, in recordings from any single

neuron), only two or three of the around a dozen voltage-dependent conductances most neurons

have, and 2) averaging data obtained from recordings of multiple neurons is invalid in systems

(like neurons) that have large numbers of fundamental components (see text).

We propose here an alternative approach in which the activity of individual neurons is

perturbed by injecting complicated patterns of current. These changes will open and close

the different voltage dependent conductances in different ways, and thus neurons with different

sets of voltage dependent conductances should show neuron-specific activities in response.

Furthermore, it is possible that only one set of voltage-dependent conductances can

reproduce these complicated responses, and thus this technique will also define voltage-

dependent make-up on an individual-neuron-by-individual-neuron basis.

Hooper Baker

2

Discussion

[Note to outside reviewers: Baker committee members are from all OU departments. This

proposal is therefore written to be understandable to non-neurobiologists. It consequently does

not have the level of detail of an NIH or NSF proposal. However, this project was supported by

an NIDA grant, and thus has been vetted by a more specialized group.]

Introduction. Nervous systems generate behavior. The nervous system transforms physical

stimuli (e.g., light, sound) into neural activity and processes this activity into sensation and

perception (e.g., color, shape, tone, rhythm). It monitors body internal state (e.g., blood glucose

level) and produces motivational and emotional drives (e.g., hunger) appropriate for each

internal state. It creates consciousness and ideation. It then decides what actions to produce.

Finally, it generates the neural signals that drive the movements that produce these acts.

Nervous systems are composed of individual cells—neurons—that communicate with each

other by generating electrical events called action potentials that propagate along the neuron’s

axon (a finger-like extension of the cell) to specialized points of contact called synapses, at

which they electrically or chemically alter the activity of their postsynaptic target neuron. One

can thus think of nervous systems as being large and complex biological computers in which the

neurons are analogous to the transistors in a silicon computer, and the axons and synapses are

analogous to the wires connecting the computer’s transistors. The fundamental goal of

neuroscience is to understand how these biological computers generate behavior.

Neurons have much more complex and individualistic input-output relationships than

transistors. For instance, in response to the same input, some neurons will fire no action

potentials, others only a few, and others a long burst of action potentials. Neurons also differ in

the absence of input, some being silent, others firing steadily, and others rhythmically firing

Hooper Baker

3

bursts of action potentials. Neurons thus have individual “personalities”. Knowing how their

neurons are interconnected is therefore insufficient for understanding how nervous systems

function. One also needs to know the intrinsic properties of the system’s individual neurons.

An analogy can be drawn by considering a business with multiple departments: floor

workers, advertising, research and development, administration. At work each department’s

personnel fulfill specific roles. Floor workers assemble the widgets the business produces in a

repetitive and soul-deadening fashion. The advertising staff generates fantastical but emotionally

compelling stories to induce people to buy things they do not need. Research and development

uses logic and experiment to make the widgets sufficiently complicated that end users cannot

repair them and competitors cannot copy them. Administrators strive to amass greater power.

Observing these people at work will not reveal their personalities because their interactions

will, presumably, alter their free expression. We could therefore also observe them during their

free time. This alone, however, may still not be enough to describe them fully: the line workers

may continue to drudge, the advertising staff to daydream, etc. To obtain a deeper understanding

of the people, we have two choices. One is to administer personality tests that measure how

much creativity, impulsivity, need for order, etc., each person has. Another is to put the people in

a reality show where they must perform arbitrary and ridiculous tasks, with the idea being that,

when forced into novel situations outside their normal activities, the various facets of their

personalities will be revealed. Importantly, if sufficiently well devised, both methods should give

the same insights, i.e., the results of the personality test should predict reality show performance

and vice versa. We propose here to finish research using an exactly analogous approach for

characterizing what makes neurons different from one another, and relating these

differences to differences in their fundamental make-ups.

Hooper Baker

4

Our preparation. We study how a small neural network, the pyloric network of the

stomatogastric nervous system (STNS), generates movements of the pylorus of the lobster

stomach. Lobster stomach muscles, like human skeletal muscles, are directly driven by motor

neuron activity. Our research is therefore relevant to understanding how nervous systems

generate movement in humans and other animals. The STNS is very advantageous

experimentally, and all the neurons of the pyloric network, and their synaptic interconnectivity,

are known (reviewed in [1, 2]). In the intact network, all the

neurons rhythmically depolarize (Fig. 1, the upward going

parts of the traces), fire a burst of action potentials, and then

hyperpolarize (the downward going parts of the traces).

However, different types of pyloric neurons produce different

activities. For instance, PD neurons fire more rapidly in the

middle of their bursts than at burst beginning and end. LP neurons fire most rapidly at burst

beginning and then continuously more slowly until burst end. PY neurons fire at a nearly

constant spike frequency throughout their bursts. The neurons also burst at different times in the

pyloric cycle: the LP neurons fire first after the PD neuron burst and the PY neurons much later.

Differing whole-cell intrinsic active properties underlie, in part, the differences in pyloric

neuron activity. Why do PD neurons rhythmically depolarize and fire bursts of action potentials?

This rhythmic activity typically occurs because another pyloric neuron, the AB neuron,

endogenously produces (i.e., self-generates) rhythmic depolarizations and action potential bursts.

The AB neuron is electrically coupled to the PD neurons, and thus passes current to them, which

drives the PD neurons to also fire rhythmically. Why do the other neurons fire after the AB/PD

neuron bursts spontaneously end? These neurons are not (under the conditions we use)

Fig. 1. Simultaneous recordings from a PD, LP, and PY neuron. All vertical scale bars 5 mV.

Hooper Baker

5

endogenously bursting neurons, and are all inhibited, not excited, by the AB/PD neurons. The

other neurons fire instead because they possess endogenous rebound properties that induce them

to depolarize and fire after inhibition (hence “rebound” firing) (Hartline et al. 1988). The

difference in when the LP and PY neurons fire is also due to differences in inherent, whole-cell

properties: LP neurons are faster rebounders than PY neurons and therefore fire first [3].

Different cell membrane conductances endow different neurons with different intrinsic

whole-cell properties [4]. Neuron electrical activity arises from the presence in neuron cell

membranes of voltage dependent ion channels. These channels open and close in response to

changes in membrane potential, and the ion flows that occur when the channels are open in turn

cause further changes in membrane potential. In the AB neuron a set of such channels allows the

neuron to rhythmically depolarize and hyperpolarize: 1) when the neuron is hyperpolarized a

background depolarizing channel is open, which causes the neuron to slowly depolarize. 2) At a

certain membrane potential, other depolarizing channels open, driving a rapid depolarization to

the burst phase. 3) This large depolarization ultimately results in hyperpolarizing channels

opening. 4) Eventually so many hyperpolarizing channels are open that the neuron begins to

hyperpolarize. 5) The depolarizing channels therefore close, causing a rapid hyperpolarization. 6)

Because the neuron is now hyperpolarized, the hyperpolarizing channels close. 7) The neuron

has now returned to state 1 and the process repeats (see [5] for a more detailed example).

The rebound properties of the other neurons similarly result from voltage dependent

conductances. The AB/PD neurons inhibit all the other network neurons. This hyperpolarization

opens depolarizing channels in these neurons. When the AB/PD neuron burst ends, these

depolarizing channels result in the inhibited neurons driving to a fully depolarized state. This

depolarization again opens hyperpolarizing currents that ultimately come to dominate the neuron,

Hooper Baker

6

end the burst, and cause the neuron to hyperpolarize. The LP neuron fires first after the PD/AB

neuron burst because its hyperpolarization-activated, depolarizing channels cause a more rapid

post-inhibitory rebound than in the other neurons [3]. The firing profiles in each neuron’s burst

similarly differ because of differences in each neuron’s voltage-dependent conductances.

The pyloric network is analogous to the business example given above. In the intact network

each neuron produces a particular activity, just as the people in each of the business’s

departments produce a certain activity. Each neuron has a “personality” defined by its intrinsic

whole-cell properties (the AB neuron’s endogenous rhythmicity; the differing rebound delays

and firing properties of the other neurons). These whole-cell properties arise from different

neurons having different ion channel make-ups, analogous to different whole-person personality

types arising from different combinations of impulsivity, emotionality, acceptance of authority,

etc. Our goal is to design a test, similar to a reality show test, that allows us to distinguish

between neurons on the basis of their intrinsic whole cell properties and, perhaps, use these

data to measure the underlying ion channel make-ups that give rise to these properties.

Hypothesis: complicated perturbation can characterize neurons sufficiently to sort

them into different classes, and perhaps to identify what sets of voltage-dependent

conductances are present in their membranes. As explained above, different pyloric neurons

produce different types of activity in the intact network. These different activities arise in part

because the neurons have different whole cell intrinsic properties…some are endogenous

bursters, others rebounding neurons with different rebound delays [1, 6, 7]. The different whole-

cell properties arise from the different neurons having different sets of voltage-dependent

channels [8]. It might thus seem there is no problem: different sets of voltage-dependent

Hooper Baker

7

conductances lead to different types of whole cell intrinsic properties which, in conjunction with

the network’s synaptic connectivity, explain how each neuron produces its characteristic activity.

However, computer modeling and further experimental work upset this tidy intellectual apple

cart. The modeling shows that many different sets of voltage-dependent conductances can lead to

essentially identical whole cell intrinsic properties and individual neuron activity [9-14]. The

experimental work shows that, even though they all produce very similar activity in the intact

network, PD (and other) neurons indeed have a wide range of voltage-dependent conductances,

with changes in one type of conductance being compensated for by changes in others [15-18].

Adding to these difficulties is the realization that average measurements of neuron

conductance make-up cannot be used to understand how neurons function [19-21]. The reasons

for this are two-fold. First, averaging only works for single-peaked populations, which the

conductance distributions may not be. A good example of how averaging fails with multi-peaked

distributions is that averaging across humans results in the average human being intersexual,

something clearly not the case. Second, even with single-peaked distributions, in systems with

many components, individuals with average values for all system components are very rare. For

instance, beautiful people are simply those whose every feature—eye spacing, ear size, chin size,

nose size, etc.—has the average value across humans [22]. It is common to have the desirable

average in one or two of these features, but to be simultaneously average in them all is very rare.

The difficulty here is that present experimental techniques allow us to measure in a single

neuron from a single individual, i.e., in a single experiment, at most two or three of the around a

dozen conductances that the neuron possesses. We thus cannot perform the equivalent of

applying a personality test to a group of humans in which we measure all the conductances in

one PD neuron, and then another, and another, etc., and then compare these conductance sets to

Hooper Baker

8

see if all PD neurons have the same conductance sets, and if they do not, how many different

types of PD neurons are there and in what particular ways they differ.

We thus have good evidence that considerable diversity exists both between different pyloric

neuron types and within neurons of a single type, but we do not have the means to sort them into

different classes, nor to state how they differ in their fundamental make-up. We hypothesized

that a “reality show” like test might separate the neurons into different classes, and allow

us to define their fundamental make-ups, individual neuron by individual neuron. The idea

here is that free-run activity of isolated neurons gives relatively little information about them,

just as observing two people’s free time is a much weaker measurement of who is most resilient,

creative, stolid, etc., than putting them into situations that challenge them.

The equivalent treatment with neurons would be to inject a wide range of different current

injections into neurons and measure the neurons’ responses. These injections will change the

neurons’ membrane potentials; these changes will open and close the neuron’s voltage dependent

conductances; the resulting ion flows across the membrane will further alter the neurons’

membrane potentials; how the neurons’ respond should therefore vary as a function of what

voltage dependent conductances the neurons possess. Further, it is possible that, if the current

injection protocols are complicated enough, that only one, or a small, set of voltage dependent

conductances would be able to reproduce the neurons’ responses, in which case these

perturbations would define what voltage dependent conductances each neuron possesses.

Prior results. This idea was the basis of a 2009-2011 NIDA award supporting computer

modeling and experimental work in my lab by Kevin Hobbs and William White. We first tested

the feasibility of the idea by constructing several computer neuron models with different sets of

conductances [23]. We then perturbed the models with complicated current injection protocols to

Hooper Baker

9

obtain a database of each neuron’s responses. These responses were then used as targets in an

evolutionary search [24, 25] in which a beginning population of model neurons with random sets

of voltage conductances evolved over time to try to produce the same set of responses. For each

model neuron the population indeed evolved

over time to find neurons that reproduced the

chosen target’s responses (Fig. 2). When the

conductances of the evolved neurons were

examined, their conductances were very similar

to those of the target neuron. This work showed that model neuron responses to complicated

current injection protocols were sufficiently distinctive that only one set of conductances could

reproduce them.

We next turned to perturbing real, isolated pyloric neurons with complicated current injection

patterns [26]. These data immediately showed that there were differences in the responses across

neurons, supporting our basic hypothesis.

When the responses of the neurons were

grouped into classes according to how similar

they were (cluster analysis [27]), it was

possible to create similarity trees, similar to

the evolutionary trees that show how closely

related different species are (Fig. 3). Cluster

analysis can be performed in several different

ways, but all trees had similar structures [26].

These data show several interesting

Fig. 2. Target activity is red; evolved model is wide green. The overlap is perfect.

Fig. 3. Similarity tree across pyloric neurons.

Hooper Baker

10

relationships. First, PD and VD neurons differ from all other neurons (A vs B). In the PD/VD

neuron group, there are two sets of PD neurons, one (C) which is most similar to VD neurons

(PD11, 16, 22) and another (D) which is not. With respect to the other neurons, IC neurons and

one set of PY neurons (PY22, 26, 7) form two closely related groups (E). The remaining PY

neurons and the LP neurons form multiple groups. One such group is composed of most of the

LP neurons (LP2, 13, 15) and a group of PY neurons (PY10, 21, 18, 3) (F). One LP neuron

(LP10) belongs to yet another PY neuron group (PY8, 15, 20) in which the similarities are less

pronounced (G, horizontal line lengths are proportional to degree of similarity).

These data suggest our hypothesis has merit, as the VD and PD neurons and PY, LP, IC, and

AB neurons also differ in that the two groups use different chemical neurotransmitters at their

synapses, and there is independent electrophysiological evidence for multiple types of PY

neurons [28]. However, two difficulties prevented us from publishing these data. First, when we

used the response waveforms as input in our evolutionary search engine, the data were unable to

drive the evolution of specific conductance sets that could reproduce the responses. Second, an

independent line of research in our lab was simultaneously showing that the fill solutions in our

electrodes (the standard one used across our field, and thus a disturbing result) was leaking out of

the electrodes and changing conductance amounts and in one case even conductance voltage

dependence [29]. We confirmed that this was occurring in our perturbation experiments by

applying the same perturbation multiple times in single experiments, which indeed showed that

the neuron responses were varying over time [26]. In this work we also found that literature

descriptions of one very important conductance in our neurons was wrong by about 5 seconds

[29], a large difference in a network that cycles with a 1 sec period.

Hooper Baker

11

Proposed research. Experiment and modeling support our hypothesis: complicated

perturbation can indeed separate pyloric neurons into distinct categories (experiment) and can

identify the conductance set individual neurons have (modeling). All the problems with our old

data have been rectified: 1) we have identified an electrode fill solution that gives stable

recordings for long periods of time even with extensive current injection, and 2) we have

corrected the old, bad description of one of the pyloric conductances [29]. We request funds for

the lobsters and supplies necessary to obtain a sufficient database for the clustering and

evolutionary fitting strategies, some 150 animals to obtain (with the inevitable failure rate that

occurs in our work) our desired set of 15 recordings each from the PD, AB, LP, and IC neurons

and 50 recordings from the PY neurons (more because of the evidence that multiple sub-classes

of PY neurons exist). We have also built a two compartment model with action potential

generation in one compartment and slow wave generation in the other. We have done so because

action potentials and slow wave voltage changes are generated in different physical locations in

the neurons [30]. Separating them in the modeling may thus help in the fitting process.

Why does this matter in the big scheme of things? The reader may well ask “Who cares how

a lobster neural network functions?” We do not study this system because of some lobster

fixation. We work in it instead because of its small size and anatomy, which make it extremely

advantageous experimentally. However, active whole cell properties and voltage dependent

conductances are present in all nervous systems [31-44]. Furthermore, how to describe and

distinguish among neurons, and to obtain complete descriptions of the voltage dependent

conductances they possess, is a general problem present across neurobiology. As such, if we

can make this technique work, it should have wide applicability across the field.

Hooper Baker

12

Bibliography

1. Nusbaum MP, Beenhakker MP (2002) A small-systems approach to motor pattern generation. Nature 417:343-350.

2. Marder E, Bucher D (2007) Understanding circuit dynamics using the stomatogastric nervous system of lobsters and crabs. Annu Rev Physiol 69:291-316.

3. Hartline DK (1979) Pattern generation in the lobster (Panulirus) stomatogastric ganglion. II. Pyloric network simulation. Biol Cybern 33:223-236.

4. Harris-Warrick RM (2002) Voltage-sensitive ion channels in rhythmic motor systems. Curr Opin Neurobiol 12:646-651.

5. Katz PS, Hooper SL (2007) Invertebrate central pattern generators. In Invertebrate Neurobiology. North G, Greenspan R (eds) Cold Spring Harbor:Cold Spring Harbor Laboratory Press.

6. Bal T, Nagy F, Moulins M (1988) The pyloric central pattern generator in crustacea: a set of conditional neuronal oscillators. J Comp Physiol A 163:715-727.

7. Hooper SL (1998) Transduction of temporal patterns by single neurons. Nat Neurosci 1:720-726.

8. Golowasch J, Buchholtz F, Epstein IR, Marder E (1992) Contribution of individual ionic currents to activity of a model stomatogastric ganglion neuron. J Neurophysiol 67:341-349.

9. Prinz AA, Billimoria CP, Marder E (2003) Alternative to hand-tuning conductance-based models: construction and analysis of databases of model neurons. J Neurophysiol 90:3998-4015.

10. Achard P, De Schutter E (2006) Complex parameter landscape for a complex neuron model. PLoS Comput Biol 2:794-804.

11. Guenay C, Edgerton JR, Jaeger D (2008) Channel density distributions explain spiking variability in the globus pallidus: a combined physiology and computer simulation database approach. J Neurosci 28:7476-7491.

12. Hudson AE, Prinz AA (2010) Conductance ratios and cellular identity. PLoS Comp Biol 6: e1000838.

13. Swensen AM, Bean BP (2005) Robustness of burst firing in dissociated Purkinje neurons with acute or long-term reductions in sodium conductance. J Neurosci 25:3509-3520.

14. Taylor AL, Hickey TJ, Prinz AA, Marder E (2006) Structure and visualization of high-dimensional conductance spaces. J Neurophysiol 96:891-905.

15. Marder E, Prinz AA (2003) Current compensation in neuronal homeostasis. Neuron 37:2-4.

16. Schulz DJ, Goaillard JM, Marder E (2006) Variable channel expression in identified single and electrically coupled neurons in different animals. Nat Neurosci 9:356-362.

Hooper Baker

13

17. Olypher AV, Calabrese RL (2007) Using constraints on neuronal activity to reveal compensatory changes in neuronal parameters. J Neurophysiol 98:3749-3758.

18. Taylor AL, Goaillard J-M, Marder E (2009) How multiple conductances determine electrophysiological properties in a multicompartment model. J Neurosci 29:5573-5586.

19. Hooper SL (2004) Variation is the spice of life. J Physiol 92:40-41 [Focus on Horn CC, Zhurov Y, Orekhova IV, Proekt A, Kupfermann I, Weiss KR, Brezina V (2004) Cycle-to-cycle variability of neuromuscular activity in Aplysia feeding behavior. J Neurophysiol 92:157-180].

20. Hooper SL, Guschlbauer C, von Twickel A, Blümel M, Hobbs KH, Büschges A (2015) Muscles: non-linear transformers of motor neuron activity. In Neuromechanical Modeling of Posture and Locomotion. Prilutsky BI, Edwards DH (eds) New York, NY:Springer.

21. Golowasch J, Goldman MS, Abbott LF, Marder E (2002) Failure of averaging in the construction of a conductance-based neuron model. J Neurophysiol 87:1129-1131.

22. Langlois JH, Roggmann LA (1990) Attractive faces are only average. Psychol Sci 1:115-121.

23. Hobbs KH, Hooper SL (2008) Using complicated, wide dynamic range, driving to develop models of single neurons in single recording sessions. J Neurophysiol 99:1871-1883.

24. Fogel DB (1994) An introduction to simulated evolutionary optimization. IEEE Trans Neural Networks 5:3-14.

25. Yao X, Liu Y, Lin G (1999) Evolutionary programming made faster. IEEE Transactions on Evolutionary Computation 3:82-102.

26. White WE (2013) Use of Empirically Optimized Perturbations for Separating and Characterizing Pyloric Neurons. PhD thesis:Ohio University.

27. Ward Jr JH (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58:236-244.

28. Hartline DK, Gassie DV, Sirchia CD (1987) PY cell types in the stomatogastric ganglion of Panulirus. In The Crustacean Stomatogastric System. Selverston AI, Moulins M (eds) Berlin:Springer-Verlag.

29. Hooper SL, Thuma JB, Guschlbauer C, Schmidt J, Büschges A (2015) Cell dialysis by sharp electrodes can cause non-physiological changes in neuron properties. J Neurophysiol 114:1255-1271.

30. Maran SK, Sieling FH, Demla K, Prinz AA, Canavier CC (2011) Responses of a bursting pacemaker to excitation reveal spatial segregation between bursting and spiking mechanisms. J Comput Neurosci 31:419-440.

Hooper Baker

14

31. Abbinanti MD, Zhong G, Harris-Warrick RM (2012) Postnatal emergence of serotonin-induced plateau potentials in commissural interneurons of the mouse spinal cord. J Neurophysiol 108:2191-2202.

32. Bennett DJ, Li Y, Siu M (2001) Plateau potentials in sacrocaudal motorneurons of chronic spinal rats, recorded in vitro. J Neurophysiol 86:1955-1971.

33. Bouhadfane M, Tazerart S, Moqrich A, Vinay L, Brocard F (2013) Sodium-mediated plateau potentials in lumbar motorneurons of neonatal rats. J Neurosci 33:15626-15641.

34. Brocard F, Verdier D, Arsenault I, Lund JP, Kolta A (2006) Emergence of intrinsic bursting in trigeminal sensory neurons parallels the acquisition of mastication in weanling rats. J Neurophysiol 96:2410-2424.

35. Brocard F, Shevtsova NA, Bouhadfane M, Tazerart S, Heinemann U, Rybak IA, Vinay L (2013) Activity-dependent changes in extracellular Ca2+ and K+ reveal pacemakers in the spinal locomotor-related network. Neuron 77:1047-1054.

36. Brocard F, Tazerart S, Vinay L (2010) Do pacemakers drive the central pattern generator for locomotion in mammals? Neuroscientist 16:139-155.

37. Butera RJ, Rinzel J, Smith JC (1999) Models of respiratory rhythm generation in the pre-Bötzinger complex. I. Bursting pacemaker neurons. J Neurophysiol 81:382-397.

38. Dekin MS, Richerson GB, Getting PA (1985) Thyrotropin-releasing hormone induces rhythmic bursting in neurons of the nucleus tractus solitarius. Science 229:67-69.

39. Hounsgaard J, Kjaerulff O (1992) Ca2+-mediated plateau potentials in a subpopulation of interneurons in the ventral horn of the turtle spinal cord. Eur J Neurosci 4:183-188.

40. Hultborn H, Zhang M, Meehan CF (2013) Control and role of plateau potential properties in the spinal cord. Curr Pharm Des 19:4357-4370.

41. Kolta A, Brocard F, Verdier D, Lund JP (2007) A review of burst generation by trigeminal main sensory neurons. Arch Oral Biol 52:325-328.

42. Martell AL, Ramirez JM, Lasky RE, Dwyer JE, Kohrman M, van Drongelen W (2012) The role of voltage dependence of the NMDA receptor in cellular and network oscillation. Eur J Neurosci 36:2121-2136.

43. Schwindt P, Crill W (1999) Mechanisms underlying burst and regular spiking evoked by dendritic depolarization in layer 5 cortical pyramidal neurons. J Neurophysiol 81:1341-1354.

44. Wang D, Grillner S, Wallén P (2013) Calcium dynamics during NMDA-induced membrane potential oscillations in lamprey spinal neurons–contribution of L-type calcium channels (CaV1.3). J Physiol 591:2509-2521.

Hooper Baker

15

Biographical Information The proposed research would be performed by Scott L. Hooper and Jeffrey B. Thuma.

Thuma will perform the experiments. Hooper will perform the cluster analyses and evolutionary searches.

Scott L. Hooper Highest academic degree: PhD, Neuroscience, Brandeis University Position at OU: Professor Duration at OU: since 1992

Other professional positions: 2006-pres, Visiting Professor (4 months/year), University of Cologne, Cologne, Germany

Refereed publications (last 5 years) EM Berg, SL Hooper, J Schmidt, A Büschges (2015) A leg-local neural mechanism mediates

the decision to search in stick insects. Curr Biol 25:2012–2017 SL Hooper, JB Thuma, C Guschlbauer, J Schmidt, A Büschges (2015) Cell dialysis by sharp

electrodes can cause non-physiological changes in neuron properties. J Neurophysiol 114:1255-1271

SL Hooper, HJ Burstein (2014) Minimization of extracellular space as a driving force in prokaryote association and the origin of eukaryotes. Biol Direct 9:24 (erratum, 2015, 10:11)

WE White, SL Hooper (2013) Contamination of current-clamp measurement of membrane capacitance by voltage-dependent phenomena. J Neurophysiol 110:257-268

JB Thuma, KH Hobbs, HJ Burstein, NS Seiter, SL Hooper (2013) Temperature sensitivity of the pyloric neuromuscular system and its modulation by dopamine. PLoS ONE 8:e67930. doi:10.1371/journal.pone.0067930

M Blümel, SL Hooper, C Guschlbauer, WE White, A Büschges (2012) Determining all parameters necessary to build Hill-type muscle models from experiments on single muscles. Biol Cybern 106:543-558

M Blümel, C Guschlbauer, SL Hooper, A Büschges (2012) Using individual-muscle specific instead of across-muscle mean data halves muscle simulation error. Biol Cybern 106:573-585

M Blümel, SL Hooper, C Guschlbauer, S Daun-Gruhn, SL Hooper, A Büschges (2012) Hill-type muscle model parameters determined from experiments on single muscles show large animal-to-animal variation. Biol Cybern 106:559-571

SL Hooper (2012) Body size and the neural control of movement. Curr Biol 22:R318-R322

Books edited or written (last five years) SL Hooper, A Büschges (eds) (in press) Neurobiology of Motor Control: Fundamental

Concepts and New Directions. New York, NY:Wiley

Hooper Baker

16

Book chapters, invited comments, book reviews, other non-refereed publications (last five years) A Büschges, SL Hooper (in press) Introduction. In Neurobiology of Motor Control:

Fundamental Concepts and New Directions. SL Hooper, A Büschges, eds. New York, NY:Wiley

SL Hooper, J Schmidt (in press) Electrophysiological recording techniques. In Neurobiology of Motor Control: Fundamental Concepts and New Directions. SL Hooper, A Büschges, eds. New York, NY:Wiley

AA Prinz, SL Hooper (in press) Computer simulation—power and peril. In Neurobiology of Motor Control: Fundamental Concepts and New Directions. SL Hooper, A Büschges, eds. New York, NY:Wiley

SL Hooper, C Guschlbauer, M Blümel, KH Hobbs, JB Thuma, A von Twickel, A Büschges (2016) Muscles: non-linear transformers of motor neuron activity. In Neuromechanical Modeling of Posture and Locomotion. B Prilutsky, ed. Springer Series in Computational Neuroscience

SL Hooper (2015) Octopus movement: push right, go left. Curr Biol 25:R366-R368 (Dispatch on G Levy, T Flash, and B Hochner (2015) Arm coordination in octopus crawling involves unique motor control strategies. Curr Biol 25:1195–1200)

SL Hooper (2015) Sensory-motor integration: more variability reduces individuality. Curr Biol 25:R991-993 (Dispatch on MJ Cullins, JP Gill, JM McManus, H Lu, KM Shaw, HJ Chiel (2015) Sensory feedback reduces individuality by increasing variability within subjects. Curr Biol 25:2672–2676)

W White, K Hobbs, SL Hooper (2014) Neuronal model output fitness function. In Encyclopedia of Computational Neuroscience. D Jaeger, R Jung, eds. New York, NY:Springer. doi: 10.1007/978-1-4614-7320-6_160-1

Jeffrey B Thuma Highest academic degree: MS, Neurobiology, Ohio University Position at OU: Research Associate/Lab Manager Duration at OU: since 1990

Other professional positions: Manager of Biological Sciences Zeiss Confocal microscope and HCOM Nikon Confocal microscope facilities.

Publications: see above

Hooper Baker

17

Other Support A. Previous University Funding. None in the last three years. We have about $400 left in Research Incentive funding left from the NIH and NSF funding SL Hooper has received over the years.

B. External Funding. Our last grant was a 2010-2013 award from the National Science Foundation, Modeling motor behavior in the lobster stomatogastric system, $399,734, which just fits within the three year window. This award helped support in part or fully the following outcomes:

SL Hooper, C Guschlbauer, M Blümel, KH Hobbs, JB Thuma, A von Twickel, A Büschges (2016) Muscles: non-linear transformers of motor neuron activity. In Neuromechanical Modeling of Posture and Locomotion. B Prilutsky, ed. Springer Series in Computational Neuroscience, pp 163-194

JB Thuma, KH Hobbs, HJ Burstein, NS Seiter, SL Hooper (2013) Temperature sensitivity of the pyloric neuromuscular system and its modulation by dopamine. PLoS ONE 8(6):e67930. doi:10.1371/journal.pone.0067930

M Blümel, SL Hooper, C Guschlbauer, WE White, A Büschges (2012) Determining all parameters necessary to build Hill-type muscle models from experiments on single muscles. Biol Cybern 106:543-558

M Blümel, C Guschlbauer, SL Hooper, A Büschges (2012) Using individual-muscle specific instead of across-muscle mean data halves muscle simulation error. Biol Cybern 106:573-585

M Blümel, SL Hooper, C Guschlbauer, S Daun-Gruhn, SL Hooper, A Büschges (2012) Hill-type muscle model parameters determined from experiments on single muscles show large animal-to-animal variation. Biol Cybern 106:559-571

SL Hooper (2012) Body size and the neural control of movement. Curr Biol 22:R318-R322

Jeff Thuma’s salary is 33% covered by Ohio University’s Neuroscience Program, 10% by HCOM, with the remaining portion being covered by other contributions that support our research. We therefore have sufficient funding to support his being able to perform this research, and need only cover the costs of lobsters and supplies for it to be performed.

Hooper Baker

18

Budget Item Amount Cost Supplier Lobsters 150 x $45 $6,750.00 Marinus Scientific Alexa Fluor 647 1 mg $251.00 Invitrogen NaCl 10 kg $86.01 Fisher KCl 1 kg $25.92 Fisher

CaCl2 1 kg $63.68 Fisher

Mg2SO4 1 kg $38.84 Fisher

Na2SO4 500 g $24.09 Fisher Tris 1 kg $124.49 Fisher Dextrose 1 kg $61.53 Fisher Maleic acid 500 g $48.30 Sigma PTX 5 g $41.60 Sigma TEA 25 g $44.70 Sigma HEPES 100 g $82.60 Sigma K Gluconate 100 g $24.70 Sigma Electrode glass 3 x $54 $162.00 WPI Grand total $7,829.46

Budget justification Lobsters: We need at least 15 good experiments for 5 of the neuron types in the ganglion (AB, LP, VD, PD, IC) to perform the cluster analysis. The PY neurons are a heterogeneous group of neurons so we want a sample size of at least 50 PY neurons. This gives a minimum of 5 x 15 + 50 = 125 animals. Isolating the LP, IC, and PY neurons requires finding and killing both PDs and the VD neuron. Isolating any of the AB, PD, or VD neurons requires finding all four of these neurons. The requested animals, 150, implies a 125/150 = 83% success rate, which may seem high given these difficulties. However, we are generally able to find both PD neurons and the VD neuron, and thus expect a high success rate for isolating LP, IC, and PY neurons. Finding all four of the AB, PD, and VD neurons is rarer. However, the experiments with isolated AB, PD, and VD experiments (3 x 15 = 45) are only 45/125 = 36% of the total experiment number, and from previous experience we expect to be able to find all four neurons in 1/3 of the experiments. We therefore believe we can obtain all the necessary data with 150 animals.

Chemicals: NaCl, KCl, CaCl2, Mg2SO4, Na2SO4, Tris base, maleic acid, and dextrose are used to make the saline that flows over the preparation during experiments. PTX (picrotoxin) and TEA (tetraethylammonium chloride) are used to block network chemical synapses. HEPES and potassium gluconate are used to make the electrode fill. Alexa Fluor 647 is a fluorescent dye used to kill cells. The amounts requested are what we expect to use in 150 experiments.

Electrode glass: This special capillary glass is used to make microelectrodes for intracellular neuron recording. The amount requested is what we expect to use in 150 experiments.

Hooper Baker

19

Recommended Reviewers 1) Astrid A. Prinz

Associate Professor (404) 727-5191 Department of Biology, Emory University, Atlanta, GA 30322 astrid.prinz@emory.edu

Dr. Prinz is the expert in stomatogastric modeling and is the one who, during her postdoctoral work with Eve Marder, most deeply investigated the ability of different conductance sets to produce near identical activity. As such, she is superbly qualified to judge this proposal. Astrid and I recently co-wrote a book chapter, but we do not collaborate on any research projects, and I therefore believe she is capable of judging the merits of this application without bias.

2) Patsy Dickinson Josiah Little Professor of Natural Sciences (207) 725-3581 Department of Biology, Bowdoin College, Brunswick, ME 04011 pdickins@bowdoin.edu

Dr. Dickinson is again an expert of long-standing in the stomatogastric field and is further qualified to appreciate the general importance of the research question as a result of her being an associate editor of the Journal of Neurophysiology. We have no research collaborations and have never published together.

3) John Simmers Researcher/Team Leader [CNRS research institutes are differently structured than US

colleges or universities. This position is a research-only position with tenure heading a lab, so Associate Professor or Professor equivalent.]

+33 05 57 57 45 60 Institut de Neurosciences cognitives et intégratives d'Aquitaine, Université Bordeaux , Zone

Nord, Bât. 2A- 2ème étage, 146 rue Léo Saignat , 33076 Bordeaux, France john.simmers@u-bordeaux2.fr

Dr. Simmers began his research career in the stomatogastric system but now studies the cellular basis of motor learning in frog (Xenopus) spinal cord. His present research focuses on the role of another electrical phenomenon, the Na-K pump that maintains the Na and K across-membrane Na and K concentration gradients, on neuron activity. He therefore has the background necessary to understand the project but also brings a point of view outside the stomatogastric system and even outside of invertebrate work. We have no research collaborations and have never published together.

Hooper Baker

20

4) Hillel Chiel Professor (216) 368-3846 Department of Biology, Case Western Reserve University, Cleveland, OH 44106 hillel.chiel@case.edu

Dr. Chiel studies motor pattern expression in the feeding system of the sea hare, Aplysia. His work combines modeling, neurophysiology, and biomechanics. Although much of his work involves muscle and the physical structures that muscles act on to produce movement, the intellectual difficulties dealt with in the grant—how to use the output of high dimensional dynamic systems to distinguish between different systems and to infer the make-up of their fundamental components—are issues that also arise in his research. As such, he is superbly qualified to judge the merits of the proposal. Again, he would also bring a non-stomatogastric point of view to the review process. We have no research collaborations and have never published together.

5) Deborah Baro Professor (404) 413-5310 Department of Biology, Georgia State University, Atlanta, GA 30302 dbaro@gsu.edu

Dr. Baro was the first researcher to use molecular biology techniques to measure voltage-dependent conductance protein expression in stomatogastric neurons. As such, she is fully aware of their variable expression in neurons of a single type and the intellectual difficulties this variation poses for these neurons nonetheless having very similar whole-cell identities and in-network activity. As such, she is again extremely well qualified to assess this proposal’s merits. We have no research collaborations and have never published together.